📄 social_science

The Mediating Effect of Deep Teaching Practices on Data Driven Quality Leadership and Deep Learning

This study validates a Quality Data-Driven Leadership (QDDL) model in Aceh Province, revealing that Deep Teaching Practices fully mediate the relationship between leadership and Deep Learning implementation, meaning leadership must first strengthen teaching practices to successfully drive educational transformation.

Original authors: Nasir Usman, Amiruddin Amiruddin, Murniati AR, Ulfah Irani Z, Niswanto Niswanto, Hendra Darmawan

Published 2026-06-24

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Nasir Usman, Amiruddin Amiruddin, Murniati AR, Ulfah Irani Z, Niswanto Niswanto, Hendra Darmawan

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Recipe for Better Learning

Imagine a school in Aceh, Indonesia, is trying to upgrade its cooking. The goal isn't just to serve food that fills stomachs (memorizing facts); the goal is to serve a gourmet meal that teaches students how to cook, think, and create for themselves (Deep Learning).

The researchers wanted to know: How do we get from the school principal's office to the student's plate?

They tested a specific "recipe" involving three main ingredients:

The Head Chef (Data-Driven Quality Leadership): The principal using data to make smart decisions.
The Cooking Technique (Deep Teaching Practices): The actual way teachers teach in the classroom.
The Meal (Deep Learning Implementation): The students actually understanding and applying what they learn.

The Main Discovery: The "Middleman" is Everything

The most surprising finding of this study is that the Head Chef (the Principal) cannot cook the meal directly.

Think of it like this: A principal can buy the best ingredients (data), write the best recipe book (strategic plans), and hire the best staff. But if the chef in the kitchen (the teacher) doesn't know how to chop, sauté, and season (Deep Teaching), the meal will still taste bad.

The study found that Deep Teaching Practices act as a "Full Mediator."

What this means: The principal's data-driven leadership only works if it first changes how the teachers teach.
The Analogy: Imagine the principal is a GPS system. The GPS can give you the perfect route (data and strategy), but the GPS cannot drive the car. The teacher is the driver. If the driver doesn't actually turn the wheel and steer (Deep Teaching), the car goes nowhere, no matter how good the GPS is.

Breaking Down the Results

1. The Principal's Role (Data-Driven Leadership)

What the paper says: Principals who use data to make decisions (like looking at test scores, student feedback, and classroom observations) are very good at improving how teachers teach.
The Result: There is a strong, positive link here. When principals use data, teachers get better at "Deep Teaching."
The Catch: The study found that the principal's data does not directly make students learn better. The link between the Principal and the Student's learning is broken unless the teacher is involved.

2. The Teacher's Role (Deep Teaching)

What the paper says: This is the most powerful ingredient. When teachers use specific methods—like asking deep questions, doing projects, and giving helpful feedback—students learn much more deeply.
The Result: This is the strongest link in the whole study. Good teaching practices are the direct engine that drives student learning.

3. The Student's Outcome (Deep Learning)

What the paper says: Students are learning to think critically, solve real-world problems, and understand concepts deeply.
The Result: This happens almost entirely because of the teachers' methods. The principal's influence is felt only because they helped the teachers get better at their jobs.

The "Full Mediation" Explained Simply

In statistics, the researchers found that the principal's direct effect on student learning was essentially zero (it wasn't statistically significant). However, the indirect effect was huge.

The Metaphor: Think of a Water Hose.
- The Principal is the water source (the tap).
- The Teacher is the hose.
- The Student is the garden.
- If you turn on the tap (Principal uses data), but the hose is kinked or broken (Teacher doesn't use Deep Teaching), the garden stays dry.
- The study proves that the water must flow through the hose to reach the garden. You cannot water the garden directly from the tap without the hose.

Who Was Studied?

The researchers looked at 175 people from five high schools in Aceh Province.

Who: A mix of Principals, Teachers, and mostly Students (since students are the ones experiencing the learning).
The Data: They used surveys (questionnaires) to ask these people how they felt about leadership, teaching methods, and learning outcomes. They used a sophisticated computer program (like a high-tech calculator) to map out these relationships.

The Bottom Line

If you want to transform a school so students learn deeply and think critically:

Don't just focus on the Principal: Having a data-savvy principal is great, but it's not enough on its own.
Focus on the Teachers: The principal's job is to use data to support and improve the teachers' classroom techniques.
The Chain Reaction: Good Data Leadership $\rightarrow$ Better Teaching Methods $\rightarrow$ Deep Student Learning.

The paper concludes that in the schools of Aceh, the path to success is clear: Leadership must fuel Teaching, and Teaching must fuel Learning. You cannot skip the middle step.

Technical Summary: The Mediating Effect of Deep Teaching Practices on Data-Driven Quality Leadership and Deep Learning

Problem Statement
The study addresses a critical gap in the implementation of the Indonesian National Curriculum Policy, which mandates a transformation toward Deep Learning (DL) to foster 21st-century competencies (critical thinking, creativity, collaboration, and communication). While the policy emphasizes deep meaning construction and knowledge application, systemic implementation in Aceh Province remains challenging. Existing literature often treats data-driven leadership and deep learning as separate domains or focuses on aggregate academic achievement rather than classroom-level pedagogical shifts. There is a lack of empirical models that explicitly link Data-Driven Quality Leadership (QDDL) to Deep Learning Implementation (DLI) through the mechanism of Deep Teaching Practices (DTP). The research seeks to determine whether school leadership influences student deep learning outcomes directly or if this influence is entirely mediated by the quality of teaching practices.

Methodology
The research employs a quantitative approach with an explanatory survey design to test causal relationships between latent variables.

Model: A Structural Equation Model (SEM) was developed where QDDL is the exogenous variable, DTP is the mediating variable, and DLI is the endogenous variable.
Data Collection: Data were gathered via a Likert-scale questionnaire administered to 175 respondents (principals, teachers, and students) across five senior high schools in Aceh Province, selected through stratified purposive random sampling.
Analysis Tool: The study utilized SmartPLS 4 software to perform Partial Least Squares-based Structural Equation Modeling (SEM-PLS).
Validation: The measurement model was validated for convergent validity (Outer Loading > 0.70, AVE > 0.50), discriminant validity (HTMT < 0.90, Fornell-Larcker criterion), and reliability (Cronbach's Alpha and Composite Reliability > 0.70). Multicollinearity was assessed using Variance Inflation Factor (VIF), and predictive relevance was evaluated using Stone-Geisser's Q² and error metrics (RMSE, MAE).

Key Results
The SEM-PLS analysis yielded the following path coefficients and statistical findings:

QDDL → DTP: Data-Driven Quality Leadership has a significant positive effect on Deep Teaching Practices ( $\beta$ = 0.542, $p$ < 0.05). This supports Hypothesis 1 (H1).
DTP → DLI: Deep Teaching Practices have a strong, significant positive effect on the Implementation of Deep Learning ( $\beta$ = 0.722, $p$ < 0.05). This supports Hypothesis 2 (H2).
QDDL → DLI (Direct): The direct effect of Data-Driven Quality Leadership on Deep Learning Implementation is statistically insignificant and slightly negative ( $\beta$ = -0.073, $p$ = 0.229). Consequently, Hypothesis 3 (H3) is rejected.
Mediation (H4): The indirect effect of QDDL on DLI through DTP is significant ( $\beta$ = 0.391). Because the direct effect is insignificant while the indirect effect is significant, Deep Teaching Practices act as a full mediator.
Model Fit and Predictive Power:
- The model explains 29.4% of the variance in DTP ( $R^2$ ) and 46.9% of the variance in DLI ( $R^2$ ).
- Predictive relevance ( $Q^2$ ) values were positive for all indicators, confirming the model's ability to predict out-of-sample data.
- The PLS-SEM model demonstrated superior predictive accuracy (lower RMSE and MAE) compared to a linear regression benchmark.
- The effect size ( $f^2$ ) of DTP on DLI is strong (0.692), whereas the direct effect of QDDL on DLI is negligible (0.007).

Key Contributions

Theoretical: The study validates a conceptual framework that positions Deep Teaching Practices as the essential bridge between leadership and learning outcomes. It challenges the assumption that leadership directly impacts student learning, demonstrating instead that leadership must first transform pedagogical practices to influence deep learning.
Empirical: It provides the first empirical evidence in the context of Aceh Province, Indonesia, confirming that without the reinforcement of Deep Teaching Practices, Data-Driven Quality Leadership does not directly translate into Deep Learning Implementation.
Methodological: The research demonstrates the utility of SEM-PLS in analyzing complex educational models with latent variables, specifically in validating a full mediation scenario where the direct path is non-significant.

Significance and Claims
The paper claims that the success of educational transformation toward Deep Learning in Aceh's high schools is highly dependent on Deep Teaching Practices. The study concludes that the role of school leadership (QDDL) is crucial as a precursor; its primary function is to facilitate, monitor, and strengthen teaching practices rather than directly influencing student learning outcomes.

The authors assert that for the Indonesian National Curriculum's mandate to be realized, school principals must utilize data not merely for administrative reporting but to drive specific pedagogical shifts. The findings suggest that investment in school data systems will only impact learning quality if that data is used to empower teachers to adopt Deep Teaching strategies (e.g., Socratic questioning, project-based learning, and constructive feedback). The study emphasizes that leadership creates the systemic conditions and professional culture necessary for teachers to adopt these practices, which in turn are the proximal drivers of student deep learning.